PH热榜 | 2026-02-16

Hey PH👋

Maor started Base44 as a side project late 2024, and it's now at millions of users.
Early on, Maor made a bet to build the entire backend infrastructure from scratch instead of relying on Firebase, Supabase, etc. That bet paid off.

Today we're releasing that same infrastructure as a standalone Backend as a Service - and it's fundamentally different from existing solutions.

The problem: If you've built with Claude Code or Cursor, you've hit this wall. The agent writes great code - then chokes on backend setup. That's not the agent's fault. Supabase, Firebase, and the rest were designed for human developers who read docs and configure things manually. AI agents don't work that way. The backend is still the bottleneck.

We built Base44 Backend from the ground up for agents:

🤖 Skills instead of APIs - AI agents don't read API docs the way humans do. Our Skills system gives agents simple, structured instructions they can execute natively. This is a huge difference in practice vs. throwing a Supabase or Firebase SDK at an agent and hoping for the best.

⚡ One command to deploy - database, auth, file storage, everything. No config files, no environment variables, no setup tax.

🔌 Built-in Connectors - OAuth-ready integrations to Gmail, Slack, Notion, Salesforce, HubSpot, LinkedIn, and more. Your agent says "post to Slack when a new order comes in" and it just works. No managing API keys or building OAuth flows from scratch.

🧠 App Agents -Built-in AI agents that live inside your app. They can read and write data, trigger backend functions, search the web, connect to WhatsApp, they're like teammates that ship with your product out of the box.

🚀 Battle-tested at scale - This isn't a new experiment. It's the same infra powering 10M+ production apps on Base44.

Who it's for: Developers using Claude Code, Cursor, or any AI coding agent who are tired of the backend being the thing that slows everything down. If you want your agent to go from idea to deployed app without you ever touching backend config, that's exactly what we built.

We're shipping fast and want to build this with the community. Tell us what's missing, what's broken, what would make this 10x more useful for your workflow.

Would love your feedback 🙏

Awesome platform!

Awesome, finally!

Is it possible to pick it up from an existing project and have two-way sync between the two? Or only start a new one?

Like the no-setup angle. npx base44@latest create fits my Cursor/Claude workflow. Curious how Skills map to real APIs + auth and what the default stack is (db, queues, logs). Any cold start quirks? Might try it on a weekend toy agent.

I’ve heard a lot about Base44!!! and got a chance to try it, it’s perfect for building quick PoC apps. I am honestly shocked with how fast and intuitive it is

Super cool.
Just took my app from localhost to deployed in a couple of minutes.
Niceeee

Been using Base44 to build internal tools and MVPs, and the speed is honestly wild. The new Backend Platform feels like it removes the biggest pain point when working with agents in Cursor or Claude Code — no setup, no config, just ship. Congrats on the launch 🚀

Fantastic, platform, have built several apps with ease, from Habits AI, to SugarQuit, VibeCall, Dialectica, Mnemo AI, and many tools for use in my own organisations too.
Always impressed with the support offered, as well as the level of growth, and speed of growth, and functionality.
Have been with them for well over 6 months now and the community and support are performance are outstanding.
Well done.
Vibe2Day

If agents like Cursor/Claude Code write the code, tools like Base44 becoming the backend makes sense remove infra, remove config, just give the agent primitives instead of APIs. That’s basically optimizing software for AI developers, not humans. Feels less like no-code and more like post code.

Amazing team and tool! Each week as I continue to build @Grapevine Software on @Base44, I think of "I wonder if this is coming to Base44" and then BOOM! it shows up as an announcement before I can even request it. Base44 over ALL OTHER AI Coder platforms!!!!

Do you pull token-level logs from each provider API, or are you reconciling against the invoice totals? That distinction matters because a single OpenAI account can have GPT-4o, embeddings, Whisper, and assistants all showing up as one line item. If ToolSpend breaks that down per model endpoint and maps it to the card charge, that's where it stops being another dashboard and starts being the thing finance actually trusts.

Wrapping auth flows, pagination, and error handling into Skills so agents don't have to parse raw REST docs is where Base44 earns its keep. The hard part is what happens when a Skill hits a partial failure or rate limit... agents retry blindly unless the surface area tells them to stop. If that stays tight, it beats wiring up Supabase from scratch every time.

That’s basically optimizing software for AI developers, not humans. Feels less like no-code and more like post code. This is fantastic. It looks like it’ll be easy to build your own and congratulations a lot

This is cool taking ideas from concept to working app instantly is powerful. 💡

For me, the real test of a tool like this is when you hit a use case that requires specialized logic (like risk scoring, fraud detection, or complex data validation).

Have you tried building something like that with a no-code tool before? If so, what was the first thing that felt like a roadblock? And if not, what would make you trust a tool enough to handle something that critical?

🔥🔥 This is going to be AWESOME for speed. Really excited to use this myself when it, hopefully, comes for existing projects with the requirement of ejecting!

Hey Product Hunt

AI tools are exploding inside companies.

What isn’t exploding? Visibility into what they actually cost.

Teams are subscribing to ChatGPT, Claude, Midjourney, Cursor, Perplexity, ElevenLabs… and finance only finds out when the bill hits.

The real problem?
AI usage (tokens) and actual spend are completely disconnected.

That’s why we built ToolSpend.

It connects your AI services + banking data and shows:
• What you're really spending
• Which teams are driving usage
• Where you’re overpaying
• Smarter model alternatives

AI shouldn’t be the next AWS surprise bill.

Excited to hear your feedback — especially from founders & dev teams already scaling with AI

Hey Product Hunt 👋

We’re the makers of ToolSpend - and we built this because we ran into the same problem ourselves.

Inside our own team, we were using ChatGPT, Claude, Midjourney, Cursor, Perplexity AI, and ElevenLabs across different projects.

Everyone was moving fast.

No one knew what we were actually spending.

Engineering saw token usage.

Finance saw card charges.

Those two worlds never met.

We realized AI spend is fundamentally different from traditional SaaS:

Usage (tokens) ≠ invoices

Teams experiment constantly

Model pricing changes fast

“Just $20/month” tools multiply quickly

So we built ToolSpend to connect AI services + banking data into one clear view:

Real AI spend across providers

Usage by team/project

Overlapping subscriptions

Smarter / cheaper model alternatives

Our goal: make AI spend observable before it becomes your next surprise bill.

We’re early and building this with founders & dev teams who are scaling fast with AI.

We’d love your honest feedback:

What’s hardest about managing AI spend today?

What metrics do you wish you had?

What would make this a no-brainer to adopt?

Thanks for checking us out 🙌

Makes sense companies suddenly have 10+ AI tools per team and nobody knows who’s using what 😅 If it actually tracks real usage + catches ghost seats before renewal, that’s legit ROI. AI spend visibility is quickly becoming finance infra, not just ops tooling.

Excited to hunt ToolSpend today 😎

Teams are rapidly adopting tools like ChatGPT, Claude, Midjourney, Cursor, and more but visibility into actual AI spend is lagging behind.

AI usage (tokens) and real cash out the door rarely live in the same place. That’s the gap ToolSpend is solving by connecting AI services with financial data to show what you’re truly spending and where you can optimize.

If you're scaling with AI, this is a problem worth paying attention to.

Congrats to the team on the launch 👏

Just when you think there isn't a tool to track all your spending accurately ToolSpend pops up amazing!

Congrats on the launch! We’ve definitely run into this with our startup team... once you’re using multiple AI models and tools, it gets challenging to understand what you’re actually spending vs. what you’re getting. Being able to chat with AI to see total spend, usage analytics, and optimization suggestions would be really valuable.

This hits home. 💸 The 'ghost licenses' problem is real and honestly, it's the same mindset I bring to building tools.

You know what else suffers from the same 'set it and forget it' drain? Risk engines. Companies pay thousands monthly for enterprise risk tools, using maybe 20% of the features, while the other 80% is just bloat they never touch.

Quick question: When you look at your stack, what's the most painful 'ghost spend' you've discovered? And if you could wave a wand what would the ideal tool look like that actually respects your budget and gives you exactly what you need, nothing more?

The number of tools I bought to test and forgot to cancel will have me bankrupt, especially browsing here everyday. this is such a smart idea!

Very Nice and useful idea, will you consider offering a white label solution as well?

Congrats on the launch! The gap between token usage and actual financial impact is very real, especially as teams experiment across multiple AI tools. How does ToolSpend attribute usage and cost to specific teams or projects when accounts and API keys are often shared across departments?

Congratulations on the launch 🎉 🎉

Cool initiative, do you also keep track of those free memberships and provide updates when the trial period is almost over?

This is super timely with everyone subscribing to 5 different AI tools. :D
Does the tool separate variable costs from fixed seat-based subscriptions?

Yep, rogue AI seats are killing us. Surprise Midjourney/Cursor/Claude charges and then finance pings me. If this actually maps tokens to teams + nudges before renewals, that’d help. Does it catch stuff paid on personal cards that get expensed later?

I think it is kinda useful for people who have like million subscription plans. :)

@visagar This is such a smart niche to go after! 👏 I (well, the whole team) use a lot of AI tools daily, and tracking what we’re actually spending manually… it gets messy fast. Love the idea!

I’m curious – how granular can the team-level visibility get? Can you attribute usage down to specific projects or cost centers, or is it mainly per tool / per team right now?

Who you gonna call? Ghost-license busters!:D 2026 is officially the year of too many AI subscriptions. Love the model alternative feature. Any plans for a one-click cancel button inside the dashboard?

This is remarkable — an open source, open weights, real-time duplex conversational AI model from NVIDIA.

Check out the demo to see how impressive this is.

So cool, I will try it to podcasting between agents in @Oasi !

Hey everyone,

ChatGPT and Claude have become a huge part of our workflows.

So we built JDoodle.ai MCP to fit directly into that flow.

JDoodle.ai MCP connects with ChatGPT or Claude or anything that supports MCP, so you can brainstorm, plan, structure, and build websites and apps inside the chat, while seeing the live preview on JDoodle.ai.

As you chat, your project gets created and updated in real time, with instant preview and deployment.

This is not just an initial generation. You can keep asking for changes, refining the UI, adding features, fixing issues, and shipping updates, all directly through ChatGPT or Claude.

Here’s an easy guide on how you can get started: https://www.jdoodle.ai/docs/mcp

Would love your feedback. We’ll be here all day answering any questions.

JDoodle AI MCP is 🔥! Migrated 3 similar no-code→fullstack (Bubble→React+Supabase). Backend scaling hurt most. Tips? Paweł @ Inigra (Clutch 5★)

the MCP integration is smart. being able to go from prompt to deployed app without leaving the chat is where vibe coding actually starts making sense

chat-driven codegen plus “publish a live link” will quickly run into state drift and unsafe execution, especially when multiple iterations touch backend, DB migrations, and secrets in the same session. Best practice: treat each chat change as a Git commit and run it in an ephemeral sandbox (devcontainers or microVMs like Firecracker or gVisor) with CI checks before deploy, and follow MCP OAuth scoping and tool allowlists for least privilege. Question: when MCP requests modify code and schema, do you generate deterministic migration plans with rollback and a diff or PR style review, or is state tracked only inside JDoodle projects?

Congrats!
I wonder if we can access to the source code of the projects which we made on the JDoodle

We built PenguinBot to go further — to actually do the work.

Give it a simple instruction, and it takes action:
• Manages emails
• Schedules tasks
• Creates documents
• Runs workflows quietly in the background

No endless prompting.
No copy-pasting between tools.
Just execution.

Our belief is simple: AI shouldn’t just assist you — it should operate with you.

PenguinBot is still evolving, and we’re building it hand-in-hand with the community.

Excited to share PenguinBot’s long-term vision :

Today, PenguinBot helps users automate workflows and connect across the tools they already use.
Our future scope is to become a true cross-platform AI workspace companion:

• Native apps on Android, iOS, Windows, and macOS for seamless access anywhere

• Smarter multi-step automations that can plan, execute, and adapt in real time

• Voice-first interactions for hands-free productivity

• Better team collaboration with shared context and sync across devices

• A growing integration ecosystem so PenguinBot fits into every workflow, not the other way around

Our goal is simple: make advanced AI automation feel effortless, personal, and available everywhere.
Would love your feedback on which feature you want us to prioritize next 🙌

Nice product. I love the design. However, I don't see any particular reason on why I should use penguinbot and not other AI orchestrating tools such as n8n or opencode. Would love to hear your opinions

If PenguinBot actually runs workflows end-to-end (not just drafts emails), that’s where value is background execution, not chat. But reliability will decide everything: autonomy is amazing… until you have to double-check every action.

I have a few questions. You mention 3000 skills, but on pricing you mention access to 30 skills. It's a bit confusing. Could you clarify?

Also, you mention an x/monitor skill. You have it on the front page. How does it work? From what I know, x restricts bots through API access.

I like the focus on actually doing things instead of just generating text. The real question is how much babysitting these agents need once they’re live. If it’s truly hands-off, that’s interesting

Congrats on the launch, turning AI from answers into actual execution is a meaningful shift. Automating emails, tasks, and workflows in the background is where real leverage starts to show up. Curious to see which integrations users ask for most and how this evolves into part of everyday work.

Voice-first is the way. We see 3x engagement with voice AI vs text chatbots. Would love to see PenguinBot add voice for customer support workflows.

long-running agents doing real tool calls will hit reliability issues fast (retries, partial failures, duplicate side effects) plus auditability for “who did what when”. Best practice: put actions behind durable execution with idempotency keys and event sourcing, e.g., Temporal or Azure Durable Functions, and add tracing plus guardrails via the OpenAI Agents SDK. Question: how are you persisting workflow state and enforcing idempotent tool execution across OAuth refresh, retries, and multi-step plans, and do you support approval gates for high-risk actions?

24/7 autonomy is powerful. Curious how you’re handling usage limits and cost control when the agent is making its own decisions — that’s going to matter a lot as these scale.

The autonomous execution angle is interesting for repetitive BD work — follow-up sequences, meeting prep, document drafts. But when client-facing communication is involved, I'd need some kind of review step before anything gets sent. How does PenguinBot handle approval workflows for outbound messages?

this is exactly what was missing. having claude code accessible from the menu bar instead of switching to terminal every time is a huge workflow improvement

The real question is: are they actually specialized… or just prompt templates with different hats? If they truly understand marketing context, that’s powerful. If not, it becomes agent overload real quick.

250+ agents is wild. As these start collaborating and triggering tools on their own, the real differentiator will be how you handle usage control and proof of execution. That layer is going to matter more than the prompts.

Hey, I noticed that all agents have the same rating and number of reviews, is that a bug?

I like how the tool shows results after typing, but I was a little bit confused by the button on the right (Is it supposed to do anything after pressing?)

Congrats on the launch! Having 250+ AI tools in one place sounds powerful, especially for busy marketers who just want to execute fast. How do you prevent overlap or tool fatigue inside Agent Squad, and help users quickly choose the right tool without getting overwhelmed by too many similar options?

Interesting approach but pricing is insane though
100$ per month is crazy for something that is open source

considering you can take a 5-10$ server and run the same thing

If OpenClaw is the engine, chowder.dev feels like the Heroku-for-claws layer — spin up, isolate, wire to channels, done. The OpenAI-compatible API angle is smart too… lowers friction hard. If it really gets you from zero → deployed agent in a minute, that’s a strong dev unlock. The agent infra stack is getting real fast.

Very cool - congrats! Where is it hosted? HOw do you think about security?

Feeling nostalgic. This reminds me of Product Hunt culture back in the day.

How do you account for different lighting conditions?

My flat white and my latte use the same milk but if they end up the same color, something went very wrong. Can I save separate presets for different drinks, or do I have to pick a favorite?

Creative idea 😀

I make cappuccino with my Bialetti every single morning, so this hits home hard lol. Getting the cream ratio right is genuinely a daily struggle — sometimes it's perfect, sometimes it's a sad beige disaster. The fact that you turned this into a real-time camera app is both hilarious and brilliant. Instant download for me.😻

nice approach...like it

Just installed Meteorite and I love it. I’ve always missed that instant notepad experience from Windows and this brings it back perfectly on macOS. Super fast access, clean UI and smooth keyboard shortcuts. It already fits into my daily workflow.

That is something that could be really helpful for me! I already payed for the pro version! I always had scattered txt files across the computer.

hi @flaskshade "Fast enough that it feels almost invisible" that's a great design north star. The difference between a 200ms and a 800ms open time is the difference between "I'll jot this down" and "I'll remember it later" (spoiler: you won't).

Curious about the Apple Notes export.. is it a one-way push, or can you pull notes back from Apple Notes into Meteorite? Bi-directional sync with Apple Notes would be a killer feature for people who want quick capture on Mac but access on iPhone later.

incredible app, congrats on the launch

Are you planning to release also an iOS version with sync between Mac and iPhone?
I would instantly make it my daily note app if that happens

Congrats on the launch! A truly frictionless capture tool in the menu bar is surprisingly hard to get right. How do you decide what not to add? With note apps, feature creep is almost inevitable, what’s your filter for keeping Meteorite minimal without users eventually asking for structure, folders, or tagging?

Hey Devs! 👋

I built MockAPI Dog to solve a problem I constantly ran into as a developer: needing APIs that don’t exist yet - or needing to test weird edge cases and errors quickly.

MockAPI Dog lets you instantly create mock REST APIs and even mock streaming LLM endpoints (like OpenAI-style or Claude-style) without creating accounts, writing backend code, or managing keys!

It’s completely free and designed to get you unblocked fast.

Would love your feedback and feature ideas! :)

PS: I'm doing zero paid promotions and no random spamming on Reddit or Twitter. I hope people upvote if they find this tool useful.

The dog branding alone got me clicking lol. I'm building a pet tech app so totally unrelated, but as a dev this is genuinely useful — mocking LLM streaming endpoints without burning API credits is a huge time saver. Nice work and congrats on the launch!🚀

Hello, solo-founder here! I'm launching this on a supposedly unlucky day, Friday the 13th, because I want to spend time with my wife on Valentine's Day tomorrow. I owe that to her.

This is my first public launch. Zero (paying) customers and zero idea if the pricing is right. But I benchmarked every competitor I could find to make sure SearchSeal is the best it could be on Day 1. Fingers crossed everything (or anything) actually works.

Try it. Break it. Tell me what sucks. I need the feedback more than I need the ego protection. Get 50% off the subscription with the promo code "LAUNCH50"!

Feel free to reach out to me at michael@searchseal.com. I read everything, mostly because no one emails me yet. Happy Valentine's Day!

edit: thanks for the info Jafar on the promo code not working

How is this different e.g. from Peec?

Hi,

My first feedback would be that the promo code is invalid. I want to explore the idea more and find out if it is a good fit for me.

Anyway, it sounds like you have a good potional on the long term 👍

Congrats on your first public launch!

Since LLMs can give different answers to the same prompt, how does SearchSeal handle this variance? Do you run the prompt multiple times and average the score to ensure the data is reliable?

Congrats! nice, It could be very useful for agencies too

Smart idea tracking AI brand mentions. Curious if you also monitor voice-based AI assistants like Siri or Alexa recommendations? Thats becoming a huge discovery channel, especially for local businesses using voice AI on their websites.

Feels like SEO is turning into “AI answer optimization.” The hard part will be proving which mentions actually drive revenue.

Congrats on the launch! I love the transparency here. Since you’re benchmarking competitors and still figuring out pricing, How are you thinking about validating pricing early on, are you optimizing for feedback volume, willingness to pay, or positioning against a specific alternative?

Hey Product Hunt! 👋

I'm thrilled to launch HookWatch today — a tool I've been building to solve a problem that's been driving me crazy for years.

The problem: Modern apps rely on webhooks, cron jobs, WebSocket connections, and now AI agents — but when they fail, they fail silently. No error page. No stack trace. Just... nothing happens. A Stripe payment webhook gets lost, a nightly backup script silently stops running, a WebSocket connection drops mid-session, an AI agent starts erroring out — and you don't find out until a customer complains or data is already gone.

I've been on the other side of that support ticket too many times. So I built HookWatch.

What is HookWatch?

It's four monitoring tools under one roof:

🔗 Webhook Monitor — Track every incoming webhook in real-time. Inspect full payloads, replay failed deliveries with one click, and get automatic retries with exponential backoff. We even have a unique request buffering feature: if your server goes down, HookWatch stores incoming webhooks and replays them when you're back online. Zero data loss.

⏰ Cron Monitor — Schedule and monitor cron jobs with human-readable syntax. Write "every day at 2am" instead of decoding 0 2 * * *. Get full execution history with stdout/stderr, automatic retries, and instant alerts when something breaks. The CLI runs jobs locally with optional cloud sync — it works 100% offline.

🌐 WebSocket Monitor — A transparent proxy for your WebSocket connections that gives you complete visibility into bidirectional traffic. Point your client to a HookWatch proxy URL instead of the original server — we forward everything transparently while capturing every message in both directions. See live connections open and close, inspect full payloads (text and binary), filter by direction (inbound/outbound), and review complete message history. Debug dropped connections, validate payload formats, and investigate incidents — all without changing a single line of your application code.

🤖 MCP Proxy — Observability for AI agent tool calls. Monitor every MCP request/response, track latency (p50/p95/p99), and get alerts when your agents hit errors. If you're building with Claude or other LLMs, this gives you the visibility you've been missing.

What makes HookWatch different?

• Four tools, one dashboard. Competitors focus on just one. We bundle webhooks + cron + WebSocket + MCP monitoring together.

• Local-first CLI. Not just a web dashboard — our CLI is a first-class citizen. Forward webhooks to localhost, run cron jobs locally, get JSON output for scripting. Works offline.

• Transparent proxying. Both for webhooks (request buffering when your server is down) and WebSockets (full bidirectional message capture). Zero code changes required.

• Human-readable schedules. Stop Googling cron syntax. Write schedules in plain English.

• AI-native from day one. MCP observability is built in, not bolted on.

I'd love to hear your feedback — what features would you want to see next? What integrations matter most to you?

Happy to answer any questions in the comments! 🚀

Webhooks are the backbone of automated billing and agent workflows. Curious how you're handling verification and tamper-proof logs once events fire — that layer’s becoming just as important as uptime.

I get uglyerror messages when trying to sign up with email, Google or Github.

Hey Product Hunt community 👋

We’re excited to share something we’ve been building for a while.

A product that answers the persistent question,”What should we fix first and why?”

There are tons of cybersecurity tools out there, but when it came time to make real decisions, something always felt off. We had plenty of signals but not enough clarity around what actually mattered versus what was just noise.

So as security engineers we built what we wished existed: a Security Decision Intelligence layer that sits above existing tools and helps teams reason about risk, not just detect it.

What Fixure does:

• Reconciles duplicated and conflicting signals into a single system reality

•  Explains downstream impact before action is taken

• Helps teams decide which actions matter most and why

• Compresses thousands of symptoms into fixable root causes

Fixure doesn’t replace scanners, it works with what you already have so teams can stop guessing at scale.

We’re opening up a beta and would genuinely love honest feedback from the community. If this resonates with your experience, we’d love to hear your thoughts and if there are other unique security products we should be learning from, please share.

Thanks for checking out Fixure!
Sameer

Interesting cybersecurity platform augmenting current tools the team is already using.

It seems like it simplifies communication, threat detection, impact analysis, facilitates decision making and reduces double efforts.

Does it also prevent hacks by assessing weaknesses before attacks happen?

Excited for this! Looks useful!

I can see how this would be helpful! Much easier than having a laundry list of tasks without any real guidance on severity.

I'm really intrigued by this, its always difficult to decide which problem is the most urget one. As someone who isnt tech savvy this will be very useful.

So when the Fixure tool helps you with cybersecurity risks, does it take information from web data. From websites such as those that are regularly updated to show more recent cybersecurity threats and attacks, along with how they are done (such as birthday attacks)?

Hey everyone! I'm Gio, the maker of DropIn.

We built this because we were frustrated with all the "missing features" in tools we use daily.

Export button in ChatGPT? - Nope. Smart replies in Slack? - Nope.

So we built DropIn. Just describe what's missing, and it gets added to any web app.

For all early adopters: we're giving free AI credits so your custom features can talk to AI directly from the browser
(summaries, smart replies, auto-responses)
Try it out: http://usedropin.com/

More integrations are on the way 😎

Would love to hear what features YOU wish your favorite tools had! 🚀

Hey Product Hunt 👋

We’ve all had that moment of staring at a tool thinking, “Why is there not just a button for this?”

Drop in is our attempt to fix that.

It’s an adaptive layer that sits on top of the web apps you already use and lets you add real features inside them. Buttons, panels, little workflow steps, even full views, just by describing what you want in plain English.

Instead of automating around tools, you change the tools themselves from the user side.

Example: you’re on a LinkedIn profile and want an “Add to HubSpot” button. You ask Drop in, it appears on the page, creates/updates the contact, logs a note the way your team wants it… and that button stays there for every profile you visit.

We’re still early, but the vision is big: let teams shape the software they already use, without waiting on vendors.

I’d love to hear: what’s the “I wish this had a button for X” moment you hit every week?

Looks very promising!

Seems powerful, congrats! So it's primarily used to instantly build integrations from your custom app to a range of different providers (e.g. Slack, ChatGPT)?

This is such a great tool! I was looking for this since long time.

Why did it took so long! 🙂

Congrats and great product. Much love.

I was not invited to the Anthropic Claude Code Hackathon (https://x.com/cerebral_valley/st... only 500 participants), but decided to participate nevertheless.

Hand vibe coded. Zero lines of code written. All Claude Code (with Zenflow UI).

Let me in!

@rajiv_ayyangar is Product Hunt becoming "pay-to-win"?

Appreciate if you're honest with your take; it might be a challenge that you're facing, and the community could help:

Some players are telling me it should have a distance counter and a leaderboard; what do you think? 🤔

@pablo_sanzo So relaxing to watch! Would be amazing with a night mode and more tracks.

It's a very original game; it's both relaxing and entertaining. I love the music, by the way!

this is actually pretty

disgustingly soothing

can you make it work for any city, not only NYC?

very fun, would love to see an update with keyboard controls

I really enjoy flying over different cities and seeing them from a bird’s-eye view. Have you thought about adding VR support? It could make the experience even more immersive!

Cool! Love the music

wow, it's a nice one — so relaxing! only drawback is that i didn't really want the plane to stop flying =))

Nice idea! I felt like the plane is going to fall, but it continue to smoothly go down 😅

Hey Product Hunt 👋
I’m Yassine, founder of PromptScan.

PromptScan helps founders and builders turn one sentence of an idea into complete market research in minutes — with competitor deep dives, customer personas, TAM/SAM/SOM, STP analysis, MVP scoping and exportable reports — all powered by real-time AI agents.

We built it because manual market research takes days/weeks — PromptScan delivers insights in under 2 minutes.

🚀 Try it for free (no credit card)

I’d love to hear what use-case you want to run first — e.g., competitive analysis, persona insights, TAM sizing? 👇

Congrats on the launch! Turning a single sentence into a structured research suite is a compelling promise, especially for early-stage founders who don’t know where to start. How does PromptScan differentiate between surface-level competitor mentions and true strategic competitors, so founders don’t base positioning decisions on noisy or loosely related players?

Market research for niche B2B segments usually means manually digging through G2, Capterra and LinkedIn for weeks. If PromptScan can shortcut that, it would be a real time saver. How well does it perform for smaller or regional markets where public data is limited?

@yassine_rahmani Congrats on the launch! 🚀 The speed and depth of this market research tool are impressive my research about "the 'Italian diaspora in London" case study it generated for me was surprisingly detailed.

I have some suggestions for improving the reports and the UX/UI. Maybe for the post-launch roadmap :) :

1. Source Citations: Since it’s an AI-driven tool, professional researchers need to verify the data (like the $14.21B market value). Adding footnotes or direct links to the web sources found by the agents would make the report more credible.

2. Visual Charts: For the 'Market Sizing' section, auto-generating a TAM/SAM/SOM chart would be a killer feature to make the report 'investor-ready' immediately.

3. UX/Accessibility areas in the 'Market Research' section that could be polished to make the reports more professional and easier to read:

   • Typography & Contrast: The font size in the main research body is a bit too small for long-form reading, and the font color is quite light. Increasing the size and using a darker, higher-contrast color would significantly improve readability.

   • SWOT Analysis UI: In the SWOT section, the 'Red on Red' color scheme makes the text very hard to decipher. A darker text color on a neutral/grey background would work much better.

   • Information Hierarchy: For sections with a lot of data, consider using dropdowns or accordions. This would allow users to collapse/expand specific fields and focus on one part of the research at a time without being overwhelmed.
     

   The product is solid and with small UI adjustments would turn a great tool into a professional-grade powerhouse. Looking forward to seeing how this evolves. Already upvoted!

Powerful tool for marketers

What sources does it use? Marketing websites? Product docs?

Congrats Yassine ! Such a great tool, I was looking for a tool like this lately to validate my idea !

Hey Product Hunt! 👋

I'm Chirag, co-maker of Jinee - an AI agent that joins your Google Meet calls and actually participates.

😩 The problem: We've all been in meetings where someone asks "what did we decide last week?" or "what's the latest ?" and the next 10 minutes are spent digging through docs, Slack threads, and memory. Meanwhile, meeting notes either don't get taken, or they're so sanitized they're useless.

🤷 Why existing tools fall short: Most meeting AI tools sit outside the call, they record, transcribe, and hand you a summary after it's over. That's useful, but it doesn't help you in the moment when you actually need answers.

✨ What Jinee does differently: Jinee joins your meeting as a participant. Say "Jinee" and ask a question; it listens, understands context, and responds out loud in real-time. Think of it as a teammate who's always prepared, never forgets, and never talks unless spoken to. 🤐

⏰ Why we built this now: Real-time speech AI finally got fast enough and good enough to make this feel natural, not gimmicky. The gap between "cool demo" and "actually useful in a real meeting" just closed.

🔑 Key benefits:

• 🎙️ Answers questions live - no more "let me get back to you on that"

• 🙌 Hands-free, voice-activated - no tab-switching or typing mid-conversation

• 💻 Works inside Google Meet - no new app for your team to adopt

• 📝 Meeting summaries delivered after the call so nothing falls through the cracks

We'd love for you to try it and tell us what breaks 🛠️ Seriously — we're early and every piece of feedback shapes what we build next.

I really like how Jinee highlights key information in real-time during meetings. Throughout my career, I’ve heard “I don’t remember that” or “we need to refine the documentation” far too many times.

That said, when thinking about enterprise adoption, data privacy becomes critical. How is sensitive meeting data handled? What guarantees are there around storage, retention, and model training?

Integrating LLMs into work pipelines is powerful, but the risk of data leakage is a real concern for companies. I’d be very interested to learn more about how Jinee approaches security and privacy at scale.

Tried it just now, the idea seems really cool. Can't wait to see how you integrate it with datawarehouses